1{"text": "{\"genre\":\"mixed_theory_empirical\",\"subfield\":\"asset_pricing\",\"subfield_secondary\":\"financial_intermediation_banking\",\"primary_method\":\"reduced_form_regression\",\"identification\":{\"estimand\":\"Association between quarterly changes in aggregate DB pension underfunding ratio (conditional on underfunded status) and changes in 30-year swap spreads\",\"strategy\":\"OLS time-series regressions of swap spread changes on changes in the underfunding ratio, with controls; 2SLS that instruments underfunding with excess stock market returns; cross-maturity and cross-country comparisons\",\"key_assumptions\":[\"Stock returns affect pension funding status through the asset side but have no direct link to swap spreads (exclusion restriction)\",\"The control variables capture other known drivers of swap spreads\",\"Dealer balance sheet constraints bind, so supply of long-dated swaps is limited\"],\"threats_addressed\":[\"The level of the yield curve jointly driving swap spreads and underfunding\",\"Omitted controls such as Treasury CDS premia and corporate bond issuance\",\"Time-period dependence of the relationship\"],\"causal_language_strength\":3,\"uses_causal_claim\":true},\"abstract_moves\":{\"question\":{\"present\":true,\"sentence_index\":1},\"gap\":{\"present\":true,\"sentence_index\":0},\"method\":{\"present\":true,\"sentence_index\":2},\"results\":{\"present\":true,\"sentence_index\":4},\"magnitude\":{\"present\":false,\"sentence_index\":-1},\"contribution\":{\"present\":true,\"sentence_index\":1}},\"introduction\":{\"moves\":[{\"paragraph_index\":0,\"move\":\"motivation\"},{\"paragraph_index\":0,\"move\":\"tension_or_puzzle\"},{\"paragraph_index\":0,\"move\":\"mechanism_or_theory_sketch\"},{\"paragraph_index\":1,\"move\":\"tension_or_puzzle\"},{\"paragraph_index\":1,\"move\":\"preview_of_approach\"},{\"paragraph_index\":2,\"move\":\"identification_defense\"},{\"paragraph_index\":2,\"move\":\"robustness_mention\"},{\"paragraph_index\":3,\"move\":\"preview_of_results\"},{\"paragraph_index\":4,\"move\":\"related_literature_positioning\"},{\"paragraph_index\":4,\"move\":\"contribution_statement\"},{\"paragraph_index\":5,\"move\":\"related_literature_positioning\"},{\"paragraph_index\":5,\"move\":\"contribution_statement\"},{\"paragraph_index\":6,\"move\":\"roadmap\"}],\"contribution_paragraph_index\":5,\"contribution_types\":[\"new_evidence\",\"new_data_or_measure\",\"new_theory_or_model\",\"resolution_of_puzzle\"],\"roadmap_present\":true,\"literature_positioning_location\":\"inside_introduction\"},\"section_roles\":[{\"idx\":0,\"role\":\"introduction\"},{\"idx\":1,\"role\":\"other\"},{\"idx\":2,\"role\":\"other\"},{\"idx\":3,\"role\":\"theory\"},{\"idx\":4,\"role\":\"results\"},{\"idx\":5,\"role\":\"results\"},{\"idx\":6,\"role\":\"conclusion\"},{\"idx\":7,\"role\":\"appendix\"},{\"idx\":8,\"role\":\"appendix\"}],\"results_presentation\":{\"economic_significance_discussed\":false,\"statistical_significance_language\":\"unclear\",\"magnitude_statements_in_intro\":false},\"robustness\":{\"location\":\"main_text_section\",\"types\":[\"instrumental_variables_like_other\",\"alternative_specification\",\"alternative_sample_period\",\"alternative_measure\",\"subsample\"].filter?[]:[\"alternative_specification\",\"alternative_sample_period\",\"alternative_measure\",\"subsample\",\"other\"]},\"tables\":{\"convention_notes\":\"Table notes define regression variables and state that the dependent variable is the quarterly change in swap spreads. Separate tables report OLS, 2SLS second-stage, cross-maturity, robustness and international results. A descriptive balance-sheet table states its units (billions of dollars) and gives its source.\",\"units_stated_in_notes\":true,\"standard_errors_description\":\"unclear\"},\"figures\":{\"uncertainty_shown\":\"no\"},\"conclusion\":{\"bounded_claims\":true,\"policy_implications\":false,\"open_questions\":false,\"restates_contribution\":true},\"hedging\":{\"level\":\"moderate\",\"typical_devices\":[\"'helps explain'\",\"'can drive'\",\"'illustrative model'\",\"'support the view'\",\"'fairly consistent evidence'\",\"'possible reasons'\"]},\"claim_evidence\":[{\"claim\":\"The 30-year US swap spread turned negative after Lehman's default and has stayed negative\",\"evidence_type\":\"figure\",\"evidence_ref\":\"Figure 1\"},{\"claim\":\"Underfunded pension plans optimally take receive-fixed positions in long-dated swaps; dealer constraints yield negative spreads\",\"evidence_type\":\"equation_or_proof\",\"evidence_ref\":\"Section III; Appendix: Model and Proofs\"},{\"claim\":\"Changes in underfunding explain changes in 30-year swap spreads when plans are underfunded, controlling for other factors\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table IV\"},{\"claim\":\"Funding status has no explanatory power for swap spreads at 2–10 year maturities\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table V\"},{\"claim\":\"The relationship survives 2SLS using stock returns as an instrument for underfunding\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table VI\"},{\"claim\":\"Results are robust to added controls, sample periods and an alternative UFR measure\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table VII\"},{\"claim\":\"Pension underfunding is linked to swap spreads in Japan and the Netherlands\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table IX\"}],\"structural_summary\":\"The paper opens by presenting a persistent pricing anomaly, negative 30-year swap spreads, and contrasting it with conventional credit-risk and convenience-yield views. It proposes a demand-driven hypothesis and documents stylized facts. Institutional background on pension duration needs and on swap supply follows. A model-free argument and a simple model link underfunding and dealer constraints to negative spreads. US regressions use a constructed underfunding measure, regime splits and 2SLS, followed by robustness tests and international replication for Japan and the Netherlands. A short conclusion restates the mechanism and the evidence.\",\"writing_profile_keywords\":[\"anomaly-driven\",\"demand-based explanation\",\"illustrative model\",\"limits to arbitrage\",\"pension funds\",\"instrumental variables\",\"international evidence\"],\"confidence\":0.75,\"notes\":\"Robustness types assigned: alternative_specification (added controls), alternative_sample_period (different time periods), alternative_measure (modified UFR), and other (the IV/2SLS test). 'subsample' is not assigned because sample splits beyond time periods are not visible. The method of reporting statistical significance is not visible in the excerpts, so it is coded unclear.\"}", "json": null, "model": "claude-opus-5-5", "stop_reason": "end_turn", "latency_s": 19.53, "usage": {"input": 7487, "output": 2347, "cache_read": 3968, "cache_write": 0}, "cost_usd": 0.077682, "request_id": "req_011CggqH1ssBDghq937Byzkm", "cached": false, "ts": "2026-11-03T23:06:27+00:00", "effort": "medium"}